I got some error from mds.log as below: 2020-07-30 18:14:38.574 7f6473346700 0 log_channel(cluster) log [WRN] : 94 slow requests, 0 included below; oldest blocked for > 619910.524984 secs 2020-07-30 18:14:43.574 7f6473346700 0 log_channel(cluster) log [WRN] : 94 slow requests, 0 included below; oldest blocked for > 619915.525079 secs 2020-07-30 18:14:44.835 7f646f33e700 -1 mds.0.159432 unhandled write error (90) Message too long, force readonly... 2020-07-30 18:14:44.835 7f646f33e700 1 mds.0.cache force file system read-only 2020-07-30 18:14:44.835 7f646f33e700 0 log_channel(cluster) log [WRN] : force file system read-only 2020-07-30 18:15:18.000 7f6473346700 0 log_channel(cluster) log [WRN] : 114 slow requests, 5 included below; oldest blocked for > 619949.950199 secs On Thu, Jul 30, 2020 at 6:55 PM Frank Yu <flyxiaoyu@xxxxxxxxx> wrote: > Hi guys, > > I have a ceph cluster with three MDS servers, two of them in active > status, while the left one is in standby-replay mode. Today I found the > message '1 MDSs are read only' show up when check the cluster status with > 'ceph -s', details as below: > > # ceph -s > cluster: > id: 3d43e9a5-50dc-4f84-9493-656bf4f06f8c > health: HEALTH_WARN > 5 clients failing to advance oldest client/flush tid > 1 MDSs are read only > 2 MDSs report slow requests > 2 MDSs behind on trimming > BlueFS spillover detected on 33 OSD(s) > > services: > mon: 3 daemons, quorum bjcpu-001,bjcpu-002,bjcpu-003 (age 3M) > mgr: bjcpu-001.xxxx.io(active, since 3M), standbys: bjcpu-003.xxxx.io, > bjcpu-002.xxxx.io > mds: cephfs:2 {0=bjcpu-003.xxxx.io=up:active,1=bjcpu-001.xxxx.io=up:active} > 1 up:standby-replay > osd: 48 osds: 48 up (since 7w), 48 in (since 7M) > > data: > pools: 3 pools, 2304 pgs > objects: 301.35M objects, 70 TiB > usage: 246 TiB used, 280 TiB / 527 TiB avail > pgs: 2295 active+clean > 9 active+clean+scrubbing+deep > > io: > client: 254 B/s rd, 44 MiB/s wr, 0 op/s rd, 15 op/s wr > > What should I do to fix the error message? it seems the cluster still > works fine(can read and write). > > Many thanks > > > -- > Regards > Frank Yu > -- Regards Frank Yu _______________________________________________ ceph-users mailing list -- ceph-users@xxxxxxx To unsubscribe send an email to ceph-users-leave@xxxxxxx